TU Berlin

Wasserwirtschaft und HydrosystemmodellierungAbstract


zur Navigation

Es gibt keine deutsche Übersetzung dieser Webseite.


The aim of this PhD thesis is to assess the long-term evolution of fine sediment budget in the impounded Upper Rhine River. With the help of long-term prognoses, it will be possible to reliably detect morphological changes in the riverbed; to estimate the impacts of induced land use changes; to apprise the sediment management strategies, e.g., to estimate residence times of contaminants, and to assess the frequency and amount of dredging required in the reservoir.

This work was developed within the framework of the joint research project "Assessments of long-term evolution of the fine sediment budgets in the Upper Rhine" between the Federal Institute for Hydrology, Koblenz, and the Technical University Berlin. The last barrage of the Upper Rhine at Iffezheim is the study area of this work. Based on the observed flow discharge and concentration values as well as the echo-sounding data for gaining riverbed bathymetry information, various methods were proposed to estimate long-term riverbed changes. The application of a high-resolution 3D model for long-term studies was the main focus in order to capture the multi-dimensional flow effects in front of the barrage as accurately as possible and to represent the local erosion and sedimentation in a realistic manner. This model serves as the basis for all further simulations.

Since practicable computation time for long-term predictions, including high-resolution 3D models for large research study areas, has not yet been considered to be satisfactory with the current technique, procedures for optimizing the calculation effort were developed during the course of the project.

Considerable upscaling approaches have to be made for the computational time-consuming numerical model. In the first step, the high-resolution model was spatially and temporal coarsened, afterwards, number of sediment fractions was reduced without affecting the quality of the results.

Using a further method, the long-term instationary simulation of riverbed volume could be replaced by a series of stationary ones. For this purpose, the boundary conditions of the numerical model–the discharge and the suspended sediment concentration–are grouped in a way that not every single event had to be simulated separately. This considerably reduced the required number of computation runs.

A prediction of future volume change in the riverbed using numerical model is only possible with projected discharge and concentration values. In order to obtain these values, time series analysis and -syntheses were carried out as well as an application of artificial neural networks.

Ultimately, the above-mentioned approaches were coupled into a new concept using their respective advantages. It was thus possible to determine a reasonable prediction of the sediment budget in the study area with a computation time, that is satisfactory to the greatest extent possible, also for the far future, the years 2070-2100.




Schnellnavigation zur Seite über Nummerneingabe